The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

COMPARATIVE ASSESSMENT OF RUNOFF AND ITS COMPONENTS IN TWO CATCHMENTS OF UPPER INDUS BASIN BY USING A SEMI DISTRIBUTED GLACIO- HYDROLOGICAL MODEL

*Syed Hammad Ali1,2, Iram Bano1, Rijan Bhakta Kayastha1, Ahuti Shrestha1

1Himalayan Cryosphere, Climate and Disaster Research Centre, Department of Environmental Science and Engineering, School of Science, Kathmandu University, Dhulikhel, Nepal ([email protected]) ([email protected]) ([email protected]) 2Glacier Monitoring Research Centre, Water & Power Development Authority, ([email protected])

KEY WORDS: Positive degree day, climate variability, glacier coverage, water resource management, River basin, basin, Upper Indus Basin

ABSTRACT:

The hydrology of Upper Indus basin is not recognized well due to the intricacies in the climate and geography, and the scarcity of data above 5000 m a.s.l where most of the precipitation falls in the form of snow. The main objective of this study is to measure the contributions of different components of runoff in Upper Indus basin. To achieve this goal, the Modified positive degree day model (MPDDM) was used to simulate the runoff and investigate its components in two catchments of Upper Indus basin, Hunza and Gilgit River basins. These two catchments were selected because of their different glacier coverage, contrasting area distribution at high altitudes and significant impact on the Upper flow. The components of runoff like snow-ice melt and rainfall-base flow were identified by the model. The simulation results show that the MPDDM shows a good agreement between observed and modeled runoff of these two catchments and the effects of snow and ice are mainly reliant on the catchment characteristics and the glaciated area. For Gilgit River basin, the largest contributor to runoff is rain-base flow, whereas large contribution of snow-ice melt observed in Hunza River basin due to its large fraction of glaciated area. This research will not only contribute to the better understanding of the impacts of climate change on the hydrological response in the Upper Indus, but will also provide guidance for the development of hydropower potential and water resources assessment in these catchments.

1. INTRODUCTION water balance method can roughly estimate the effects of glacier and snow in monthly or larger time scale. Pakistan is situated in South Asia between 24°-37°N latitude Approaches that relate glacier melt water production (gained by and 66°-77°E. It hosts the triple point (junction) of three world the measurements or modelling) with measured discharge famous mountain rages Himalayas, and Hindukush further downstream are problematic because glacier melt water in its north. There are more than 5000 glaciers feeding the Indus can be considered as raw volume input into the runoff system, from 10 sub-basins through different tributaries ranging from but the discharge further downstream has been modified by few tens of meters to more than 70 km long. A major proportion precipitation, evaporation, irrigation, damming, or exchange of flow in the Indus River is contributed by its snow and with subsurface flow regimes and groundwater (Kaser et al., glacier-fed river catchments situated in the Karakoram Range. 2010). The isotopic investigation cannot be used extensively as The major Indus basin is divided into three basins the Kabul, it requires large financial and laboratory support. However, the Upper Indus, and Panjnad. The Upper Indus basin includes the application of hydrological models to understand the glacier Gilgit, Hunza, Shigar, Shyok, Zanskar, Shingo, Astor, and effects in hydrology is comparatively new and more commonly Upper Indus sub-basins (Bajracharya and Shrestha, 2011). used (Hagg et al., 2007; Huss et al., 2008; Koboltschnig et al., These basins feature distinct hydrological regimes, which are 2008; Prasch, 2010; Nepal et al., 2013). But the main problem linked with the main source (snow and glacier) of their melt- is the availability of long term data of high quality to symbolize water generation and can be differentiated by its strong the hydrological dynamics of Upper Indus Basin (UIB). correlation with the climatic variables. The hydro-climatic conditions in glacierized basins of the UIB The hydrological system of Upper Indus basin is mainly reliant alter considerably with elevation and topography and the on the monsoon and melting of snow and glaciers. To know the behavior of glaciers to climate changes may be dissimilar at contribution of glaciers to runoff is an essential step to identify higher elevation than lower elevation, predominantly for large the impact of climate change on water resources, flooding and glaciers where thick/thin debris cover can suppress/increase the drought in glacier fed basins. At present there are four melting rate (Hewitt, 2005, 2011; Kaab et al., 2012; Gardelle et approaches to investigate the runoff components from rainfall, al., 2012). melting of snow and glaciers: the water balance analysis According to the findings of Archer, 2003; Fowler and Archer, (Thayyen et al., 2005; Kumar et al., 2007), glacier degradation 2005 and Hasson et al., 2015, the initial water supply from the from observation or modelling as contribution to runoff UIB after a long dry period (October to March) is gained from (Kotliakov, 1996; Kaser et al., 2010), isotopic investigations melting of snow (late-May to late-July), the range of which (Dahlke et al., 2013) and hydrological modelling (Hagg et al., largely depends upon the accumulated snow amount and 2007; Naz et al., 2013) though limitations of these methods concurrent temperatures and snowmelt runoff is then cannot be ignored, especially for climate change studies. The overlapped by the glacier melt runoff (late-June to late-August),

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1487 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

the extent of which primarily be contingent to the melt season elevation ranges from 1432 - 7849 m a.s.l. Shuttle Radar temperatures. Topography Mission Digital Elevation Model (SRTM DEM Climate change is explicit and increasingly serious concern due 2000) of resolution 90m of U.S. National Aeronautics and to its recent acceleration globally. Glaciers and icy surfaces are Space Administration (NASA) is used to delineate the the most sensitive indicators of global warming which have catchment boundaries of the study area. The total numbers of shown their immediate response in terms of mass balance and glaciers in basin are approximately 1,384. The clean glacier area contribution of melt water to the sea level rise. According to is 3,673.04 km2 and the debris cover area is 479.56 km2. The World Meteorological Organization statement on status of snow cover area in the Hunza River basin varies from climate, the first decade (2001-2010) is the warmest decade approximately 80% in winter to 30% in summer (Tahir, et al., recorded over the globe and 2010 ranked as the warmest year 2011). The distribution of area versus elevation for both basins (+0.53°C) followed by 2005 (+0.52°C) and 1998 (0.52°C). is shown in Figures 2 and 3. Sixteen warmest years of the globe occurred during the last two decades. Similarly, local impacts of the regionally varying climate change can differ significantly, depending upon the local adaptive capacity, exposure and resilience (Salik et al., 2015), especially for the sectors of water, food and energy security. In view of high sensitivity of mountainous environments to climate change and the role of melt water as an important regulator for UIB runoff dynamics, it is very important to study the snow and glacier and their impacts on the hydrologic regime of this Upper Indus region under climate variability to manage the available water resources. The aim of this study is to understand the glacier and snow effects on hydrological regime in Upper Indus basin and to identify the contributions to runoff components Figure 2. Hypsograph of Hunza River Basin 2. STUDY AREA The Gilgit River basin with drainage area, 13,471 km2 (Fig. 1), encompasses eastern part of the Hindukush Range and drains southeastward into the Indus River. Gilgit River is measured at Alam Bridge hydrometric station. Geographically the basin extends from 35.80°N, 72.53°E to 36.91°N, 74.70°E. The elevation of the basin ranges from 1,250 - 7,730 m a.s.l. Approximately 982 km² of catchment area is at an elevation above 5000 m and almost the same area (8%) is glaciated accounting for 4% of the UIB cryospheric extent. The clean glacier area is ≈944 km2 and the debris cover area is ≈146 km2. The average SCA varies from approximately 85% in winter to 10% in summer (Tahir, et al., 2011). Gilgit River basin receives its precipitation from both westerly disturbances and summer Figure 3. Hypsograph of Gilgit River Basin monsoon system. In Gilgit River basin there are 585 glaciers and 605 glacier lakes, whereas, 8 potentially dangerous glacier 3. HYDROCLIMATIC DATA ANALYSIS lakes. Hydro-meteorological characteristics of each river have been determined through water yield calculation and basin analysis. Temperature and precipitation varies place to place due to altitudinal variation, monsoon path and topography. The Indus River and its tributaries rise in the sparsely populated glaciated mountains of western and central Asia. The Indus River itself contributes more than half the total flow and has a controlling storage at Tarbela Dam as the river emerges from the mountains (Hayley et al., 2005). The UIB perceives contrasting hydro-meteorological regimes mainly because of the complex terrain of the HKH ranges and sophisticated interaction of prevailing regional circulations (Hasson et al., 2014a, 2015). The sparse (high and low altitude) meteorological network in such a difficult area neither covers fully its vertical nor its horizontal extents it may also be highly influenced by the complex terrain features and variability of the Figure 1. Study Area meteorological events. The hydrological regimes of Gilgit and The Hunza River basin with drainage area of 13,713 km2 (Fig. Hunza as sub catchments of UIB due to its complex terrain, 1), geographically extends from 36.05°N, 74.04°E to 37.08°N, highly concentrated cryosphere and the form, magnitude and 75.77°E. It is situated in the high-altitude central Karakoram seasonality of moisture input associated with two distinct modes region, with a mean catchment elevation of 4,631 m. of prevailing large scale circulation; westerly disturbances and Approximately 4,152 km2 of catchment area is glaciated. The summer monsoon and river runoff is mainly dependent on these

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1488 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

circulations. Two thirds of the precipitation in the Karakoram while precipitation shows increasing trend in both the stations region is a result of westerly’s influences and one third have a with 0.3 mm yr-1 and 11.41 mm yr-1 respectively for the same strong monsoon component (Hewitt, 2009). time period (Fig. 6 and 7). WAPDA and PMD has maintained some automatic weather stations equipped with precipitation gauges as shown in (Fig. 1) at Gilgit, Yasin and Ushkor in Gilgit and Naltar, Ziarat and Khunjerab in Hunza basin located between elevation 1460 – 4700 m a.s.l. The data from these stations are used for this study. Some year’s precipitation data is missing in Yasin and Ushkor stations due to some problem in the gauges. Mean annual precipitation and temperature at different stations of both basins are presented in Table 1. Variations in temperature and precipitation at different climate stations are demonstrated in Figures 4 & 5 and 8 & 9 respectively. The decrease in precipitation at higher altitude is partly due to wind induced error especially in the case of snowfall where losses can be on average 10–50%, regardless of the many gauges furnished with wind shields Sevruk (1985, 1989) and Forland et al. (1996). Figure 6. Annual Temperature trend at climate stations of both basins

Figure 7. Annual Precipitation trend at climate stations of both Figure 4. Variations in temperature at different climate stations basins (Hunza River Basin)

Figure 8. Variations in precipitation at different climate stations Figure 5. Variations in temperature at different climate stations (Hunza River Basin) (Gilgit River Basin)

Naltar is located in the south of the Hunza River basin and has the highest precipitation and the strongest influence of the monsoon (June, July, August [JJA]). The JJA variability is also the highest in Naltar, underlining the strong variability in monsoon strength. A non-parametric test Mann Kendall test which is a pragmatic choice that has been extensively adopted for hydro-climatic trend analysis has been done for studying the temporal trends of hydro-climatic series which shows no statistically significant trend in both temperature and precipitation data with p > 0.05 value. From this data analysis, it is observed that temperature is slightly decreasing at Naltar climate station (Hunza River basin) by -0.0097°C yr-1 on the other hand Yasin climate station (Gilgit basin) show an Figure 9. Variations in precipitation at different climate stations increasing trend by 0.036°C yr-1 over a period of 1995 -2013 (Gilgit River Basin)

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1489 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

According to the available data, mean annual runoff from Gilgit and Hunza rivers gauged at Alam bridge and Daniyour bridge, where, M is the snow or ice melts (mm/d), T is the air respectively are around 300 m3/s (1995 - 2013), and 323 m3/s temperature (ᵒC) and DDF is the positive degree day factor for (1966 - 2013) respectively, which is a large portion of the UIB snow or ice (mm/d/°C). mean annual runoff. Discharge is low from the month of A transition scheme outlined by Kayastha et al., 2000 is used to January to April, start to pick up in May with the advent of separate snow and rain in this study. summer and reach higher levels of discharge during July and August, continuing up to September, and then start to drop from October before the starting of winter and again become low during the months of November and December. The highest The snow and ice melt in precipitation contributing to runoff monthly average flow is in July at both discharge stations and the base from each elevation zone is calculated as given in Fig.10. The discharge regime can be classified as glacial with equation 2. The runoff from all zones is summed to get the maximum runoff in high flow period May to October and non- runoff from the entire basin (Q) as given in equation 3. The glacial with minimum runoff during the low flow period flow chat of MPDDM is shown in Fig. 3. November to April. Qz = Qr x Cr + Qs x Cs + Qb (2)

(3) 3 where,QZ is the discharge (m /s) from zone Z, Qr and Qs are the discharge (m3/s) from direct precipitation and snow and ice melt, respectively, C is runoff coefficient with Cr referring to rain and Cs to snow and ice melt as mentioned in Martinec 3 (1975) and Qb is the base flow (m /s). The discharge Q is then routed to the basin outlet as per the recession equation 4 given by Martinec (1975).

(4) 3 where, Qn is the river discharge (m /s) at the basin outlet on nth day and k is the recession coefficient. Figure 10. Monthly average flow at discharge stations of both 5. HYDROLOGIC SIMULATION basins The MPDDM was set up on two catchments of Upper Indus basin, Hunza and Gilgit river basins. Using a single criterion in 4. METHODS the calibration and validation processes constrains the model The Modified Positive Degree Day Model (MPDDM) version parameters to fit certain characteristics of the observed data and 1.0 has been developed at the Himalayan Cryosphere, Climate neglects the remaining features. We adopt multiple criteria to and Disaster Research Center (HiCCDRC), Kathmandu assess the model performance: the Nash-Sutcliffe Efficiency University, Nepal in 2014. The MPDDM is based on the (NSE), the relative mean error (RME) and volume difference relation that the melting of snow or ice during any particular (VD) (Table 3). NSE is sensitive to high peaks (Nash & period is proportional to the positive degree-days linked by the Sutcliffe, 1970); RME is for the relative error; and VD positive degree day factor involving a simplification of complex represents the overall difference between observation and process that are more properly described by the energy balance simulation. Therefore, the combination of these three criteria of the glacier surface and overlaying atmospheric boundary can give a proper assessment of the model performance in layer (Braithwaite and Olesen, 1989). This approach is runoff simulation. appropriate in region with scarce data as it requires less input data and uses simple equation to estimate melt. In this model, 6. MODEL CALIBRATION AND VALIDATION the both basins are divided in to 33 elevation zones with 200 m bands. The model is forced by a combination of local 6.1 Hunza River Basin meteorological observations. The local meteorological forcing Model calibration involves constraining model parameters to data required by the model include air temperature, obtain the best fit between the simulated results and the precipitation, at daily time step. For each elevation band, the air available observed data. These parameters are adjusted within temperature and precipitation time series are interpolated ranges based on previously published values to minimize the according to its mean elevation using altitude-dependent lapse difference between simulated and observed daily runoff for the rates. The air temperature and precipitation are assumed to period 2000–2013. The integrated model performance for decrease with increasing altitude which was derived from simulating runoff is assessed by comparing simulated with observed temperature and precipitation data at the different observed daily runoff for both calibration (2000–2004) and stations (Table 2). Daily runoff data from the gauging stations validation (2008–2013) periods. In validation years three years located at the river outlet of the catchments (Fig. 1) are used to (2005-2007) discharge data is missing, therefore we have not evaluate the model results. In each zone, daily snow and ice used these periods for validation. The results of calibration and melt from the glacerized and glacier free areas is calculated validation periods for Hunza River basin are shown in Table 4. using equation 1. Graphically, simulated versus observed daily runoff are shown M= {DDFSnow/Ice/Debris x T if T > 0 in Figures 11 and 12. The average observed discharge was 291.51m3/s and 316.10m3/s and simulated discharge is 0 if T < 0 (1)

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1490 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

279.59m3/s and 272.51m3/s for calibration and validation period, respectively.

Figure 14. Calibration of MPDDM Gilgit River Basin

7. RUNOFF COMPONENTS Figure 11. Calibration of MPDDM Hunza River Basin We calculated variations in runoff and its components in the both catchments of UIB using observed meteorological data. In Hunza River basin a large fraction of runoff is from snow-ice melt 38.73%. Irrespective of glacier runoff, base flow contributes 50.84% of total runoff, rainfall runoff is 10.43%. On the other hand the contributions of different components of runoff like snow-ice melt, base flow and rainfall in Gilgit River basin are 26.58%, 61.11% and 0.12% respectively. Base flow is the technical name for the dry weather flow and much of the wet weather flow in a stream or river. River base flow results Figure 12. Validation of MPDDM Hunza River Basin from ground water seeping into riverbanks or the riverbed. The flow may be significant enough to allow the stream to flow year 6.2 Gilgit River Basin round (i.e., perennial or permanent stream). Without base flow The model was calibrated from 1999 to 2005, and validated recharge from ground water to streams and rivers, many would from 2006 to 2010 at a daily time step. The results of not carry a flow of water except during storms. Streams that calibration and validation periods for Gilgit River basin are flow only periodically in response to rainstorms or seasonal shown in Table 5. For graphical representation, the observed snowmelt events are known as ephemeral or intermittent and simulated discharge in calibration and validation period is streams (American Ground Water Trust, 2003). In an ideal presented in Figures 13 and 14. The average observed discharge basin the geologic material are of uniform permeability and the was 298.18m3/s and 316.60m3/s and simulated discharge is steam has excellent hydraulic continuity with the nearby aquifer 286.99m3/s and 304.07m3/s for calibration and validation period rarely if ever is this situation come across in nature most all respectively. basins have complex geology and streams draining such basins Calibration and validation results clearly shows that the model may have incomplete hydraulic continuity with the underlying can efficiently simulate hydrologic response of the Hunza and aquifers in many cases more than one aquifer may be Gilgit River basins and give a good fit of the low flow as well as contributing water to the streams (Ferris et al., 1962). high flow using the optimal parameter sets (Table 6), with a According to genetic components of stream flow for the Hindu rating of ‘very good’ on the scale of Moriasi et al. (2007), Kush Karakoram, Pamir and Tien Shan mountains established except some sudden peaks due to some extreme events like by (Dreyer et al., 1982) rivers are fed by groundwater from (cloud burst, GLOF, eruption of ponds and etc.) during November to February. Snowmelt begins in the lower parts of monsoon period which is common in hydrological modeling but basin in March, and groundwater discharge increase to the uncertainty of the results cannot be ignored. The sources of maximum in July – August, until September Rivers are fed by the uncertainty are data used in the calibration and validation. groundwater, ice melt, snowmelt and precipitation. Snowmelt Among the observed data of precipitation, temperature and ceases by October in high mountains and rivers are fed solely discharge, the precipitation data are expected to be the largest by groundwater. The method is based on regular changes in source of uncertainty because most stations are located at lower stream flow structure during the year. elevations, while information about precipitation at higher Monthly variation in the magnitude of the contribution of each elevations (above 5000 m a.s.l) is missing. Additionally in this runoff component to the total runoff and its intra-annual study, a degree-day method for snow melting was used and the distribution are presented for the period 1999–2013 in Figures snow cover area is not updated or modelled. By doing this, 15 and 16. According to our estimation the mean annual glacier there is unlimited snow available for melting, which might not runoff in Hunza and Gilgit River basin are observed between realistic in a long term simulation and may lead to high May - October 47.71% and 33.2%, respectively, the fraction of estimation of water from rain-base flow as compared to snow- glacier runoff approaches zero during the winter months, and ice melt. runoff from the glacier-free zone of the catchment is a larger component of total runoff. With increasing temperature, the contribution from glacier runoff begins to increase and reaches maximum in May (≈65%) and is also significant during the other summer months (≈40 – 55%) in Hunza River basin. On the other hand glacier runoff is maximum in May (≈50%) and ≈20 – 45% in other summer months in Gilgit River basin. Rainfall is the largest contributor in July and August. The melting of glaciers along with the more precipitation fall as rain is likely to cause heavy flooding and Hunza and Gilgit River

Figure 13. Calibration of MPDDM Gilgit River Basin

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1491 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

basin is susceptible to floods in monsoon as recorded in 2005 Table 3. Model performance assessment criteria and their and 2010. corresponding formulation Criteria Formula Value Range Perfect Value

NSE 푛 (푂 − 푆 )2 -∞,1 1 1 − 푖=1 푖 푖 푛 2 푖=1(푂푖 − 푂푖)

VD 푉 − 푉′ Least volume difference 푅 푅 ∗ 100 푉 indicate the good fit of 푅 model

푛 RME 푖=1(푆푖 − 푂푖) -∞,+∞ 0 푛 푖=1 푂푖

Where Oi and Si are the observed and simulated flow, respectively; i is the time series index; n is the total number of time steps; VR is measured runoff volume; VR′ is simulated runoff volume. Figure 15. Monthly contribution of runoff components from

Hunza River Basin Table 4. Model efficiency of calibration and validation periods

on Hunza River Basin Period NSE VD RME

Calibration 0.88 4.09% -0.04

Validation 0.89 13.79% -0.13

Table 5. Model efficiency of calibration and validation periods on Gilgit River Basin Period NSE VD RME

Calibration 0.85 3.75% -0.04

Validation 0.78 3.96% -0.04

Figure 16. Monthly contribution of runoff components from

Gilgit River Basin Table 6. List of calibrated parameters used in MPDDM

Table.1 Mean annual precipitation and temperature at different Parameters Description Values stations of both basins ks Degree day factor for snow ablation 5.0-6.0 mm/°C/day (Jan – Apr) Station Elevation Period Mean Annual Mean Annual (m a.s.l) Temperature (°C) Precipitation (mm) (Bocchiola, et al., 2011) 6.0-8.0 mm/°C/day (May – Sep) Gilgit 1460 1995 – 2013 16.0 155 6.5-7.0 mm/°C/day (Oct – Dec) Naltar 2898 1995 – 2013 6.0 669 ki Degree day factor for ice ablation 6.0-8.0 mm/°C/day (Jan – Apr)

Ushkor 3051 1995 – 2013 6.0 394 (Bocchiola, et al., 2011) 6.0-9.0 mm/°C/day (May – Sep)

Yasin 3280 1995 – 2013 5.0 364 7.5-8.5 mm/°C/day (Oct – Dec)

Ziarat 3668 1995 – 2013 3.0 285 kd Degree day factor for debris covered 4 - 6 mm/°C/day ice (Mihalcea, et al., 2006) Khunjerab 4730 1995 – 2013 -5.0 209 Γ Temperature Lapse Rate 0.48 -0.76 °C/100m

PG Precipitation Gradient 0.25 -1 Table 2. Details of climatic and Discharge stations, their Cr Coefficient of Rain 0.08-0.5

coordinates, data period, and sources of data. Coefficient of Snow. 0.1-0.40 Cs Sr. Station Basin Elevation (m) Coordinates Period of Source (Tahir, et al., 2011) No. Record Lat Long X Constant for Recession Coefficient 0.97

Climate Stations Y 0.04 1 Naltar 2898 36.168 74.175 1995-2013 WAPDA

2 Ziarat Hunza 3668 36.829 74.418 1995-2013 WAPDA 3 Khunjerab 4730 36.812 75.332 1995-2013 WAPDA 8. CONCLUSION 4 Yasin 3280 36.451 73.294 1995-2013 WAPDA

5 Ushkor Gilgit 3051 36.027 73.415 1995-2013 WAPDA The mountains region of UIB is a key source of water for

6 Gilgit 1460 35.921 74.327 2000-2013 PMD Pakistan and delivers the main water source for the IBIS, one of

Discharge Stations the world’s largest assimilated irrigation networks. The River Indus is nourished by an amalgamation of melt water from 1 Daniyour Br. Hunza 1450 35.56 74.23 1960-2013 WAPDA seasonal and permanent snow fields and glaciers, and runoff 2 Alam Br. Gilgit 1430 35.56 74.19 1995-2013 WAPDA from rainfall both during the winter and monsoon season. To

simulate the runoff and its contribution components in Upper

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1492 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

Indus, the MPDDM was used to identify the runoff proportions Ferris, J.G., and others, 1962, theory of aquifer tests: U. S. from rain and base flow, snow-ice melt in two catchments of Geol. Survey Water suppl Paper 1936E, p.99-101. Upper Indus basin: Hunza River basin, and Gilgit River basin. The results show that the MPDDM is able to give rational Forland, E.J. et al., 1996. Manual for operational correction of assessment of runoff on these two catchments. The hydrological Nordic precipitation data.24/96, DNMI, P.O. Box 43, Blindern, components differ between both basins and these affect the flow Oslo, Norway. of Indus River and the water resource use in the lower Indus. For the Gilgit river basin, the largest contributor to runoff is Fowler, H. J. and Archer, D. R.: Hydro-climatological rain and base flow, whereas melting of snow-ice is markedly variability in the Upper Indus Basin and implications for water dominant in the Hunza River basin. The variation of resources, in: IAHS Publ. 295, Regional Hydrological Impacts hydrological components within the basin was due to the of Climatic Change – Impact Assessment and Decision Making, elevation range and the distribution of areas within each Proceedings of symposium S6, Seventh IAHS Scientific elevation band in the catchment, temperature variations, Assembly, Foz do Iguaçu, Brazil, 2005 permanent snowfields and the glacierized proportion. The more precipitation that falls as rain is likely to raise the high flow and Gardelle, J., Berthier, E., Arnaud, Y., and Kääb, A.: Region- glacier retreating is likely to reduce the base flow, which wide glacier mass balances over the Pamir-Karakoram- possibly leads to more droughts and floods. In summary, these Himalaya during 1999–2011, The Cryosphere, 7, 1263–1286, two study basins are vulnerable to climate change. The model doi:10.5194/tc-7-1263-2013, 2013. performances can be enhanced by using some other global climate data and emissions scenarios to assess the potential Hagg, W., Braun, L. N., Kuhn, M. & Nesgaard, T. I. (2007) impacts of climate change on the hydrology of whole Upper Modelling of hydrological response to climate change in Indus basin as a macro scale model. glacierized Central Asian catchments. Journal of Hydrology

332(1–2), 40–53. ACKNOWLEDGEMENTS Hasson, S., Lucarini, V., Pascale, S., and Böhner, J.: We acknowledge support from the Contribution to High Asia Seasonality of the hydrological cycle in major South and Runoff from Ice and Snow (CHARIS) Project, University of Southeast Asian river basins as simulated by PCMDI/CMIP3 Colorado, Boulder, USA funded by the United States Agency experiments, Earth Syst. Dynam., 5, 67–87, doi:10.5194/esd-5- for International Development (USAID).We thank the project 67-2014, 2014a. partners for their help in carrying out this study, including: the Himalayan Cryosphere, Climate and Disaster Research Center (HiCCDRC), the Department of Environmental Science and Hasson, S., Pascale, S., Lucarini, V., and Böhner, J. (2015): Engineering, School of Science, Kathmandu University, Glacier Seasonal cycle of precipitation over Major River Basins in Monitoring Research Center (GMRC); Water & Power South and Southeast Asia: a review of the CMIP5 climate models data for present climate and future climate projections, Development Authority (WAPDA) Government of Pakistan. J. Atmos. Res.

REFERENCES Hayley J. Fowler and David R. Archer,” Hydro-climatological variability in the Upper Indus Basin and implications for water Archer, D.R., (2003). Contrasting hydrological regimes in the resources”, Regional Hydrological Impacts of Climatic Change- Upper Indus Basin. J. Hydrol. 274, 198–210. Impact Assessment and Decision Making ,Proceedings of symposium S6 held during the Seventh IAHS Scientific American Ground Water Trust (2003). Groundwater and river Assembly at Foz do Iguaçu, Brazil, April 2005, IAHS Publ. 295 flow. THE AMERICAN WELL OWNER, 2003, Number 3 (2005).

Bajracharya, SR; Shrestha, B.,(eds) 2011. The status of Glaciers Hewitt, K.: The Karakoram anomaly? Glacier expansion and the in Hindu Kush-Himalayan region. Kathmandu: ICIMOD. “elevation effect”, Karakoram Himalaya, Mt. Res. Dev., 25, 332–340, 2005. Braithwaite, R. J., and Olesen, O.B., (1989): Calculation of glacier ablation from air temperature, West Greenland. In Hewitt, K (2009). Glacier change and Himalayan cryosphere. Glacier fluctuations and climate change. Ed. Oerleamas, J., Lecture presented at global change impact study center. Kluwer Academic Publishers, Dordrecht, 219-233. Islamabad, Pakistan. Dahlke, H. E., Lyon, S. W., Jansson, P., Karlin, T. & Rosqvist, Hewitt, K.: Glacier change, concentration, and elevation e_ects G. (2013) Isotopic investigation of runoff generation in a in the Karakoram Himalaya, Upper Indus Basin, Mt. Res. Dev., glacierized catchment in northern Sweden. Hydrological 31, 188–200, doi:10.1659/MRD-JOURNAL-D-11-00020.1, Processes. doi: 10.1002/hyp.9668. 2011. Dreyer, N. N., Nikolayeva, G. M., and Tsigelnaya, I.D., (1982) Hussain, M., Farinotti, D., Bauder, A. & Funk, M. (2008) Maps of streamflow resources of some high mountain areas in Modelling runoff from highly glacierized alpine drainage basins Asia and North America Hydrological Aspects of Alpine and in a changing climate. Hydrological Processes 22(19), 3888– High Mountain Areas (Proceedings of the Exeter Symposium, 3902. July 1982). IAHS Publ. no. 138.

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1493 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W7, 2017 ISPRS Geospatial Week 2017, 18–22 September 2017, Wuhan, China

Kaab, A., Berthier, E., Nuth, C., Gardelle, J., Arnaud, Y. Sevruk, B., 1989. Reliability of precipitation measurement, Contrasting patterns of early twenty-first-century glacier mass International Workshop on Precipitation Measurement. WMO change in the Himalayas. Nature 488, 495-498 (2012). Tech. Document, pp. 13–19.

Kaser, G., Großhauser, M. & Marzeion, B. (2010) Contribution Tahir, A.A., Chevallier, P., Arnaud, Y., Ahmad, B., 2011. Snow potential of glaciers to water availability in different climate cover dynamics and hydrological regime of the Hunza River regimes. In: National Academy of Sciences of the United States basin, Karakoram Range, Northern Pakistan. Hydrology and of America (ed. by R. G. Barry), 20223–20227. Earth System Sciences 15(7):2275–2290. www.hydrolearth- syst-sci.net/15/2275/2011/; accessed 4 August Kayastha, R. B., Ageta, Y., and Nakawo, M., (2000). Positive 2011.http://dx.doi.org/10.5194/hess-15-2275-2011. Degree-day factors for ablation on glaciers in the Napalese Himalayas: case study on Glacier AX010 in Shorong Himal Thayyen, R. J., Gergan, J. T. & Dobhal, D. P. (2005) Nepal. Bulletin of glaciological research, 17. 1-10. Monsoonal control on glacier discharge and hydrograph characteristics, a case study of Dokriani Glacier, Garhwal Koboltschnig, G. R., Schöner, W., Zappa, M., Kroisleitner, C. Himalaya, India. Journal of Hydrology 306(1–4), 37–49. & Holzmann, H. (2008) Runoff modelling of the glacierized

Alpine Upper Salzach basin (Austria): Multi‐criteria result validation. Hydrological Processes 22(19), 3950–3964.

Kotliakov, V. M. (1996) Variations of Snow and Ice in the past and at present on a Global and Regional Scale. Citeseer.

Kumar, V., Singh, P. & Singh, V. (2007) Snow and glacier melt contribution in the at Pandoh Dam, Himachal Pradesh, India. Hydrological Sciences Journal 52(2), 376–388.

Martinec, J., (1975). Snowmelt-runoff model for stream flow forecasts. Nordic Hydrology, 6(3), 145-154. Doi:10.2166/nh. 1975.010.

Moriasi D.N., Arnold, J., Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L., 2007. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans. Am. Soc. Agric. Biol. Eng., 50(3), 885–900

Nash, J. E. & Sutcliffe, J. V. (1970) River flow forecasting through conceptual models part I—A discussion of principles. Journal of Hydrology 10(3), 282–290.

Naz, B., Frans, C., Clarke, G., Burns, P. & Lettenmaier, D. (2013) Modeling the effect of glacier recession on streamflow response using a coupled glacio-hydrological model. Hydrology and Earth System Sciences Discussions 10(4), 5013–5056.

Nepal, S., Krause, P., Flügel, W.A., Fink, M. & Fischer, C. (2013) Understanding the hydrological system dynamics of a glaciated alpine catchment in the Himalayan region using the J2000 hydrological model. Hydrological Processes doi: 10.1002/hyp.9627.

Prasch, M. (2010) Distributed process oriented modelling of the future impact of glacier melt water on runoff in the Lhasa River Basin in Tibet. Doctor, Ludwig-Maximilians-University of Munich, Germany.

Salik, K. M., Jahangir, S., Zahdi, W. Z., and Hasson, S. (2015.): Climate change vulnerability and adaptation options for the coastal communities of Pakistan, Ocean Coast. Manage.

Sevruk, B., 1985. Correction of precipitation measurements, Workshop on the Correction of Precipitation Measurements. WMO/IAHS/ETH, Zurich, Switzerland, pp. 13–23.

This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLII-2-W7-1487-2017 | © Authors 2017. CC BY 4.0 License. 1494